Applying MDL in PSO for learning Bayesian networks

Shu Ching Kuo, Hung Jen Wang, Hsiao Yi Wei, Chih Chuan Chen, Sheng Tun Li

研究成果: Conference contribution

1 引文 (Scopus)

摘要

Since learning Bayesian networks from data is difficult, a new approach is proposed. The particle swarm optimization (PSO) and minimum description length (MDL) are combined to obtain a suitable Bayesian network. MDL is the fitness function in this learning algorithm to evaluate the goodness of the network. By adopting MDL, the balance between simplicity and accuracy is assured, which enables the optimal solution for complex models to be found in reasonable time. Base on the MDL principle, the PSO is used to enhance the structure learning in Bayesian networks. Moreover, conditional probabilities associated with the Bayesian networks are then statistically derived from these data. In the end, the Stroke data set is used for testing the efficiency and effectiveness of the stable network. Experimental results show that the proposed approach has a good accuracy than the comparative methods.

原文English
主出版物標題FUZZ 2011 - 2011 IEEE International Conference on Fuzzy Systems - Proceedings
頁面1587-1592
頁數6
DOIs
出版狀態Published - 2011 九月 27
事件2011 IEEE International Conference on Fuzzy Systems, FUZZ 2011 - Taipei, Taiwan
持續時間: 2011 六月 272011 六月 30

出版系列

名字IEEE International Conference on Fuzzy Systems
ISSN(列印)1098-7584

Other

Other2011 IEEE International Conference on Fuzzy Systems, FUZZ 2011
國家Taiwan
城市Taipei
期間11-06-2711-06-30

指紋

Bayesian networks
Bayesian Networks
Particle swarm optimization (PSO)
Particle Swarm Optimization
Structure Learning
Conditional probability
Fitness Function
Stroke
Learning algorithms
Learning Algorithm
Simplicity
Optimal Solution
Testing
Learning
Evaluate
Experimental Results
Model

All Science Journal Classification (ASJC) codes

  • Software
  • Theoretical Computer Science
  • Artificial Intelligence
  • Applied Mathematics

引用此文

Kuo, S. C., Wang, H. J., Wei, H. Y., Chen, C. C., & Li, S. T. (2011). Applying MDL in PSO for learning Bayesian networks. 於 FUZZ 2011 - 2011 IEEE International Conference on Fuzzy Systems - Proceedings (頁 1587-1592). [6007570] (IEEE International Conference on Fuzzy Systems). https://doi.org/10.1109/FUZZY.2011.6007570
Kuo, Shu Ching ; Wang, Hung Jen ; Wei, Hsiao Yi ; Chen, Chih Chuan ; Li, Sheng Tun. / Applying MDL in PSO for learning Bayesian networks. FUZZ 2011 - 2011 IEEE International Conference on Fuzzy Systems - Proceedings. 2011. 頁 1587-1592 (IEEE International Conference on Fuzzy Systems).
@inproceedings{2f883d94004844b9afdb17906e924e62,
title = "Applying MDL in PSO for learning Bayesian networks",
abstract = "Since learning Bayesian networks from data is difficult, a new approach is proposed. The particle swarm optimization (PSO) and minimum description length (MDL) are combined to obtain a suitable Bayesian network. MDL is the fitness function in this learning algorithm to evaluate the goodness of the network. By adopting MDL, the balance between simplicity and accuracy is assured, which enables the optimal solution for complex models to be found in reasonable time. Base on the MDL principle, the PSO is used to enhance the structure learning in Bayesian networks. Moreover, conditional probabilities associated with the Bayesian networks are then statistically derived from these data. In the end, the Stroke data set is used for testing the efficiency and effectiveness of the stable network. Experimental results show that the proposed approach has a good accuracy than the comparative methods.",
author = "Kuo, {Shu Ching} and Wang, {Hung Jen} and Wei, {Hsiao Yi} and Chen, {Chih Chuan} and Li, {Sheng Tun}",
year = "2011",
month = "9",
day = "27",
doi = "10.1109/FUZZY.2011.6007570",
language = "English",
isbn = "9781424473175",
series = "IEEE International Conference on Fuzzy Systems",
pages = "1587--1592",
booktitle = "FUZZ 2011 - 2011 IEEE International Conference on Fuzzy Systems - Proceedings",

}

Kuo, SC, Wang, HJ, Wei, HY, Chen, CC & Li, ST 2011, Applying MDL in PSO for learning Bayesian networks. 於 FUZZ 2011 - 2011 IEEE International Conference on Fuzzy Systems - Proceedings., 6007570, IEEE International Conference on Fuzzy Systems, 頁 1587-1592, 2011 IEEE International Conference on Fuzzy Systems, FUZZ 2011, Taipei, Taiwan, 11-06-27. https://doi.org/10.1109/FUZZY.2011.6007570

Applying MDL in PSO for learning Bayesian networks. / Kuo, Shu Ching; Wang, Hung Jen; Wei, Hsiao Yi; Chen, Chih Chuan; Li, Sheng Tun.

FUZZ 2011 - 2011 IEEE International Conference on Fuzzy Systems - Proceedings. 2011. p. 1587-1592 6007570 (IEEE International Conference on Fuzzy Systems).

研究成果: Conference contribution

TY - GEN

T1 - Applying MDL in PSO for learning Bayesian networks

AU - Kuo, Shu Ching

AU - Wang, Hung Jen

AU - Wei, Hsiao Yi

AU - Chen, Chih Chuan

AU - Li, Sheng Tun

PY - 2011/9/27

Y1 - 2011/9/27

N2 - Since learning Bayesian networks from data is difficult, a new approach is proposed. The particle swarm optimization (PSO) and minimum description length (MDL) are combined to obtain a suitable Bayesian network. MDL is the fitness function in this learning algorithm to evaluate the goodness of the network. By adopting MDL, the balance between simplicity and accuracy is assured, which enables the optimal solution for complex models to be found in reasonable time. Base on the MDL principle, the PSO is used to enhance the structure learning in Bayesian networks. Moreover, conditional probabilities associated with the Bayesian networks are then statistically derived from these data. In the end, the Stroke data set is used for testing the efficiency and effectiveness of the stable network. Experimental results show that the proposed approach has a good accuracy than the comparative methods.

AB - Since learning Bayesian networks from data is difficult, a new approach is proposed. The particle swarm optimization (PSO) and minimum description length (MDL) are combined to obtain a suitable Bayesian network. MDL is the fitness function in this learning algorithm to evaluate the goodness of the network. By adopting MDL, the balance between simplicity and accuracy is assured, which enables the optimal solution for complex models to be found in reasonable time. Base on the MDL principle, the PSO is used to enhance the structure learning in Bayesian networks. Moreover, conditional probabilities associated with the Bayesian networks are then statistically derived from these data. In the end, the Stroke data set is used for testing the efficiency and effectiveness of the stable network. Experimental results show that the proposed approach has a good accuracy than the comparative methods.

UR - http://www.scopus.com/inward/record.url?scp=80053087663&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80053087663&partnerID=8YFLogxK

U2 - 10.1109/FUZZY.2011.6007570

DO - 10.1109/FUZZY.2011.6007570

M3 - Conference contribution

AN - SCOPUS:80053087663

SN - 9781424473175

T3 - IEEE International Conference on Fuzzy Systems

SP - 1587

EP - 1592

BT - FUZZ 2011 - 2011 IEEE International Conference on Fuzzy Systems - Proceedings

ER -

Kuo SC, Wang HJ, Wei HY, Chen CC, Li ST. Applying MDL in PSO for learning Bayesian networks. 於 FUZZ 2011 - 2011 IEEE International Conference on Fuzzy Systems - Proceedings. 2011. p. 1587-1592. 6007570. (IEEE International Conference on Fuzzy Systems). https://doi.org/10.1109/FUZZY.2011.6007570